Is the GOFORDEV Index a valid measure?

Tracking Good Governance and Local Development: Is the GOFORDEV Index a valid measure? Joseph J. Capuno*, Ma. Melody S. Garcia** and Janette S. Sardalla**

1. Introduction How well does the Good Governance for Local Development (GOFORDEV) Index track the quality of local governance? This is the main question answered in this paper, which partially evaluates the sensitivity of the Index, as a whole, to its component indicators and its consistency with other similar indices using primary data.

An empirical evaluation has now become necessary, given the myriad of indicators already developed, which all presumably have been vetted for their theoretical merits. The 2001 compilation of indicators of good governance prepared by the Philippine Business for Social Progress (PBSP) already contains 30 sets initiated by academic institutions, government agencies, civil society organizations and donor agencies. The list does not even include other previously adopted initiatives such as the DSWD’s Minimum Basic Needs (MBN) and the HDN-UNDP’s Human Development Index (HDI).

Surely, however, the profusion of indicators systems is partly justified given the complexity of governance issues confronting the Philippines, and the various approaches taken toward these issues, which, in turn, then lead to diverse policy proposals and action programs for different stakeholders. Nonetheless, PBSP’s attempt at classification should be taken a step further towards consolidation and cooperation among the concerned proponents, if only to re-channel the *

Assistant Professor of Economics at the University of the Philippines and Project Director of the Governance Project of the Philippine Center for Policy Studies (PCPS). ** Research Associate, PCPS Governance Project.

1

considerable resources spent on measurements to actual, useful social intervention programs.

Approaches at consolidation can be both theoretical and empirical. But where the indicators systems all pass theoretical evaluation, a resort to empirical tests will then help achieve the objective. Many of those in the PBSP’s list have been empirically verified, although at various levels. This paper presents the partial results of an attempt to validate the GOFORDEV Index, developed by the Philippine Center for Policy Studies (PCPS) under its Governance Project1. The results are based on the pilot test of the Index in twelve cities and municipalities in Bulacan and Davao del Norte in 2001. In summary, the GOFORDEV Index offers some suggestions to improve the quality of local governance, is found not unduly biased by any of its component indicators, and also generally consistent with a few similar indicators.

The rest of the paper is arranged as follows: The conceptual framework behind the GOFORDEV Index is described in Section 2, followed by a discussion in Section 3 of the methodology used in evaluating the Index. The results are presented in Section 4. Concluding remarks end the paper.

2. The conceptual basis of the GOFORDEV Index2

Many Filipinos are living in abject poverty, manifested in terms of high prevalence of malnutrition, shorter life expectancy, poor health and education status, substandard living conditions and limited economic opportunities [World 1

This is a multi-phased project whose main objectives are to develop a set of simple, valid and useful index of good governance and local development and to advocate the wider use and eventual institutionalization of the index. Developed in 1999-2000, the GOFORDEV Index is pilottested for two years (2001-02) in Bulacan and Davao del Norte. The advocacy phase of the project is scheduled in 2003. 2 Substantial portions of this section are lifted from Capuno [2000], which contains a more extensive discussion of the conceptual foundation of the GOFORDEV Index and its similarities and differences with some existing indicators.

2

Bank 2000]. And, judging by the reports on the causes of the persistent Mindanao crisis, political and human rights abuses and social discrimination (or indifference) compound the poor’s economic miseries. Unsurprisingly, and rightly so, poverty reduction, or more generally economic development, therefore remains a major policy objective [e.g., see Medium Term Philippine Development

Plan 1999-2004]. But since poverty is more directly felt at the local level, its alleviation necessarily must be a local government priority. Appropriately, therefore, local governments should be evaluated in terms of their impact on local development and welfare, the improvement of which is their raison d’etre. A cursory review of

local development plans would suggest that this is so. Although local development and welfare result from a host of factors, many of which are beyond the control of a local government unit (LGU), it nonetheless has vast fiscal and other resources to promote local conditions either directly or indirectly. Oftentimes, however, the size of its fiscal resources is not enough; how these are generated and used can prove to be the more critical factor. Fiscal decisions for example are likely to lead to more favorable results if the various stakeholders in the area are also involved in the decision process, and not only in the evaluation of fiscal outcomes. The participation of the target beneficiaries, civil society organizations, local business sector and other interest groups is therefore important; some would even argue that democratic participation is the. sine qua non of governance. With their involvement, the people can thereby direct their local governments to concentrate on the provision of basic public services in a more transparent and accountable manner.

Hence, a measure of good local governance should be based on the local government’s impact in terms of improved development and welfare, as

3

evaluated by its constituents, and in the manner by which it is able to elicit the people’s active participation in local public affairs. Basically, therefore, the linkages between public service provision, development impact and people’s participation may be illustrated using the diagram below (Figure 1). The constituents demand public services from their local government, which in turn provides the required public services within its powers and means. Under ideal conditions, a perfect matching of needs and provision occurs. But since real conditions are far from ideal, the people’s direct participation in public decision-making processes becomes necessary. Aside from its intrinsic value, participation also minimizes the risk or costs of erroneous local government decisions due to lack of information, corruption, indifference or sheer incompetence. With the people’s direct involvement, local governments therefore become more transparent, accountable, fair, effective and efficient.

Figure 1. The conceptual framework of the GOFORDEV Index

Public service provision (DOI)

Local Government

Participation (PDI)

Constituency

Consultation

Public service needs (DNI)

The diagram guides the development of the GOFORDEV Index, which is a composite of three sub-indices, each of which comprises several indicators (Table 1). Although simpler, the GOFORDEV Index is akin to the Governance

4

Quality Index in scope and construction [Manasan et al. 1999]. To the extent that it promotes local development and therefore local poverty alleviation programs, it shares a similar objective with the Good Governance Indicators for Poverty Programme Assessment introduced by the Ateneo School of Government and UNDP [Brillantes et al. 2000]. The GOFORDEV sub-indices are also each related to some narrower-focused indicators systems.

The first sub-index is the Development Needs Index (DNI) which indicates how well the local constituents regard the performance of their local government in terms of providing basic public services (e.g., day care services, health services and water supply), in solving local community problems and in helping them improve their overall family condition. To rate the local government’s achievements in these aspects, five indicators are included in the DNI. The DNI score thus reflects the proportion of the local population that positively appraises their local government in terms of these five aspects of good governance. The DNI is similar to the MBN index and the World Bank’s Filipino Report Card survey [World Bank 2001] in that the quality of local governance is measured from the people’s perspective.

In contrast to the DNI, the Development Orientation Index (DOI) takes the point of view of the local government, an approach shared with the DILG’s LPPMS and the Local Government Capability Classification Scheme proposed by the Local Government Development Foundation. Based on widely available public documents (such as fiscal data), the DOI reflects the local government’s relative preference for social services and economic services, which include health, education, housing, social welfare, livelihood programs and infrastructures [see Table 1 for the component indicator of the DOI]. In a sense, therefore, the DOI gauges the local government’s social priority that help promotes human development [see Philippine Human Development Report

1997].

5

6

Table 1. The GOFORDEV Index and its Component Indicators Index GOFORDEV Index

Objective To measure the quality of local governance

Development Needs Index (DNI)

To measure relative access to and adequacy of basic public services

Indicators/Sub-Index Development Needs Index (DNI), Development Orientation Index (DOI) and Participatory Development Index (PDI) Adequacy of Health Services Ratio

Adequacy of Day Care Services Ratio

Access to Sources of Drinking Water Ratio

Development Orientation Index (DOI) Participatory Development Index (PDI)

Formula = 1/3 [DNI + DOI + PDI]

Health Service = [(Number of respondents who are aware that there is rural health center with a regular doctor in their barangays)/(total number of respondents)] x 100 Day Care = [(Number of respondents who are aware that there is a day care center with a regular teacher or a social worker in their barangays) /(total number of respondents)] x 100 Drinking Water = 100- [(Number of respondents who reported difficulty in getting drinking water)/(total number of respondents)] x 100]

To measure the perceived effectiveness at solving public problems To measure the perceived effect on family condition in life

Addressing Public Problem Ratio

To measure people’s assessment of local government’s performance To measure the relative prioritization for developmentoriented public services

(see above) Development Expenditure Priorities

Development Expenditure Priorities = [(Expenditures on social services and economic services*)/(Total expenditures*)] x 100 *net of personnel services (Note that DOI is equal to Development Expenditure Priorities)

To measure the degree of people’s direct participation in local consultative or decision-making bodies

Participation in Local (Municipal/City) Development Council

Local Development Council = 100 if the Municipal/City Development Council was convened at least twice last year with the presence of private sector representative(s); 0 if not School Board = 100 if the Local School Board was convened at least once last year with the presence of the president or representative of the Parents-Teachers Association; 0 if not Barangay Consultation = [(Number of respondents who were consulted at least once last year by their mayor, vice mayor, barangay officials or members of the Sangguniang Bayan)/(Total number of respondents)] x 100 Barangay Meeting = [(Number of respondents who reported that the mayor or the vice mayor attended at least one public meeting in their barangay last yea)/(Total number of respondents)] x 100 = ¼ (Local Development Council + School Board + Barangay Consultation + Barangay Meeting)

To measure the degree of public consultations

Effect on Family Condition Ratio

Participation in Local School Boards

Barangay-Level Consultation Ratio (BLCR)

Attendance in Barangay Meetings Ratio

To measure the extent of democratic governance through participation and consultation

(see above)

7

Public Problem Ratio = (Number of respondents who are aware of some pressing public problems and report that the local government addresses these problems)/(total number of respondents)] x100 Family Condition Ratio = (Number of families who report that their local governments contributed positively to the improvement of their condition in life)/ (total number of families) x100 = 1/5 [EFCR + EAPPR + ADCSR + AHSR + ASDWR]

It is widely known that oftentimes the level and quality of public services provided are not what are exactly needed, thus necessitating a mechanism that will bridge the gap between public service provision and needs. One such mechanism is the people’s participation in local planning activities and their consultation by local officials. At least four formal venues for a wider and active participation including the Local School Board, Local Health Board and Local Development Councils are mandated under the Local Government Code of 1991. The people’s participation in some of these venues and the extent to which local officials conduct barangay-level consultations are captured by the third subindex, the Participatory Development Index (PDI). Based on both household surveys and available public documents, the PDI therefore is similar to the attempts of CODE-NGO and other groups at measuring democratic governance under the LGC of 1991.

Like its component sub-indices and indicators, the GOFORDEV Index has a range of score from zero to 100. Therefore, a score close to 100 implies the locality’s proximity to “good governance”, although a score far from 100 does not imply “bad governance”. Thus, the GOFORDEV Index, like GNP growth, exchange rate and inflation, aims to provide a quick diagnosis of what ails a locality, the details of and the solution to which should be discussed in local consultative bodies and in other occasions.

3. An empirical validation

Overview This section briefly describes the pilot test design adopted to empirically validate the GOFORDEV Index.3 The Index is pilot-tested in twelve cities and

3

A more detailed description of the pilot test design and its implementation is contained in Capuno, Garcia and Sardalla [2001].

8

municipalities of Bulacan and Davao del Norte. The two pilot test provinces contrast well in terms of human and socioeconomic development, with Bulacan having a relatively higher score in Human Development Index and greater fiscal revenues. The contrast therefore helps isolate critical factors to achieving good governance other than fiscal resources and initial levels of development, at least at the province level.

The randomly-selected pilot-test areas are likewise grouped in terms of their relative levels of socioeconomic development. In Bulacan, the high-income LGUs are San Jose del Monte City, Baliuag and Plaridel, while the low-income localities are Guiguinto, Angat and Bustos. In Davao del Norte, on the other hand, the relatively well off areas are Panabo City, Sto. Tomas and Tagum City, while the less well off areas are Braulio Dujali, Island Garden City of Samal and Asuncion. In these areas, two major activities are undertaken: the generation of the locality’s score in the GOFORDEV Index through data collection and processing (including a random survey of 100 households), and the dissemination of the scores through public presentations and the publication and distribution of

komiks and posters.4 The main responsibility over these activities is delegated to local area partners who are contracted by the PCPS for the purpose. The local area partners are the local government units themselves (in San Jose del Monte, Guiguinto, Panabo City and Braulio Dujali), business groups and civic organizations (in Baliuag and Angat), academic institutions (in Plaridel, Bustos, Tagum) and NGOs (in Sto. Tomas, Asuncion and Samal City). In these activities, the pilot-test areas are provided by the PCPS with financial, technical and logistical support. In addition, the local agents are monitored and evaluated in terms of competence, credibility and effectiveness. In its entirety, the pilot test therefore should not only lead to a refined GOFORDEV Index, but yield some

4

While it does not measure transparency and accountability per se, the GOFORDEV Index however provides the occasion where such aspects of governance may be promoted.

9

guidelines as well for the installation of the Index in other areas and by other groups.

As argued above, an empirical evaluation has now become necessary to help consolidate the number of existing indicators. Such examination can be undertaken at different levels, three of which are employed to test the GOFORDEV Index. At the lower level, a Delphi method is adopted in which a group of experts are consulted and asked to evaluate the Index. A number of focus group discussions, public consultations in and out of Metro Manila and interviews have all been undertaken under the PCPS Governance Project. In addition, four background papers and four baseline studies of provinces in the country were previously commissioned under the PCPS Governance Project to provide inputs to the development of the GOFORDEV Index.5

At the upper-level is of course an indication of whether the proposed index yields the desired result in terms of improved governance (processes and procedures) or development outcomes (local welfare or conditions). An assessment of the impact of the GOFORDEV Index on local fiscal decisions and processes and on local welfare (e.g., citizen’s satisfaction) is the main motivation behind the pilot-test phase of the Governance Project.

At the intermediate level, the one undertaken in this paper, actual data are used to check the sensitivity of the index to its component indicators, and its consistency with other similar indicators. These tests are partially employed to identify some of the internal and external flaws of the Index.

5

These are Bautista and Juan [2000], Burton [2000], Estrella [2000], Lim, Doñgail and Oliveros [2000], Pingol and Alonzon [2000], Rocamora and Villamil [2000], Solon, Fabella and Capuno [2001] and Veneracion [2000].

10

Data issues and methodology Since the local area partners are the ones tasked to collect and process data, a host of technical and motivational problems could arise with regard to the validity and consistency of the information generated, which are fed into the GOFORDEV index to obtain the localities’ scores. To ensure the technical soundness of the data collected and processed, the PCPS developed a fairly detailed sampling design for each of the pilot test areas, the questionnaire that was used in the household survey, and the data processing software to compute the GOFORDEV Index. In addition, the PCPS trained, supervised and monitored the agents in the conduct of surveys and processing of the data.

As a final countercheck against the results reported by the agents, an independent household survey was conducted using the same sampling design as the one the adopted by the agents.6 Hence, a comparison of the two sets of data collected in the same areas should suggest the reliability, competence and effectiveness of the agents, who may be motivated by other considerations aside from meeting its contractual obligations with the PCPS. For example, the agent may rig the scores, despite the guidelines set by the PCPS, for political ends. The existence of such rigging may therefore be checked with the independentlygenerated household-level data.

Sensitivity test and correlation analysis To establish the consistency of the GOFORDEV Index, one sensitivity test and a correlation analysis are employed. The sensitivity test shows how the relative rankings of the pilot test areas change with the modifications in the weights applied to the component indicators or sub-indices. A fairly stable ranking will lend some confidence to the present formula of the GOFORDEV

6

The baseline household surveys in Bulacan and Davao del Norte were collected with the Bulacan State University-Malolos Campus and the Ateneo de Davao University, respectively.

11

Index. Furthermore, it would indicate that none of the component indicators, which may be grossly measured (due to lack of data, carelessness, incompetence or malice), unduly biases the overall result. With such a property, the scores in the GOFORDEV Index may then still be comparable if a poorly calculated indicator is excluded in the computation of the overall score.

The correlation analysis, on the other hand, depicts the observed relationship between the GOFORDEV Index and four other indicators of good governance. The four other indicators are constructed here using the same data set contained in the baseline household survey, thus minimizing the noise that could arise from the differences in the sampling design, survey instruments and motivations of two data collection activities. A positive correlation therefore is expected between the GOFORDEV Index and each of the following indicators defined below:

1.

Membership in local organization ratio – this is the percentage share of the total number of respondents who are reported to be a member of any organization in his or her community, to the total sample of households. Thus, the ratio indicates the “sense of community” among the respondents in the area.

2.

Trust in elected officials ratio – this is the percentage share of the total number of respondents who are reported to trust their elected officials, to the total sample of households. In this case, the ratio suggests the quality of local leadership and the relationship between the government and the governed population.

3.

Net satisfaction rating of the local mayor – this is the number of respondents who are reported to be satisfied with their mayor’s performance minus the number of those who reported otherwise. This indicator is akin to the ratings of national government officials regularly published by the Social Weather Station.

12

4.

Desirable changes in public services ratio – this is the percentage share of the total number of respondents who noted desirable changes in local public service provision, to the total sample.

4. Analysis of results7

To provide further background to the analysis, the respective scores of the pilot test areas in the GOFORDEV Index, as reported by the local agents, are shown in Table 2. Two major observations can be gleaned from the table. First, the scores in the GOFORDEV Index of the pilot areas in Davao del Norte are generally higher than those in Bulacan. This observation remains valid even if the outliers in each province are taken out of the picture.

Additionally, the relatively highly developed areas in Bulacan have higher scores in the GOFORDEV Index than their less developed counterparts in the province. The opposite case is true however in the case of Davao del Norte where the relatively poor areas appear to compensate for their lower socioeconomic status with a quality of governance better than their more prosperous neighbors. The results therefore imply that the GOFORDEV Index might be capturing some aspects of good governance that are quite independent of the locality’s level of economic development – a finding similar to the inexact relationship commonly observed between the overall quality of governance and economic development of some countries.

One aspect of governance where the pilot test areas in Davao del Norte particularly stand out is the extent of democratic participation. People’s participation and consultation in public affairs seem to be wider in Davao del Norte than in Bulacan, as indicated by the two provinces’ average scores in PDI which are 57 and 44.5, respectively. Arguably, the wider participation in Davao del Norte has led to higher allocation to basic public services, as manifested by 7

Substantial portions of this section are lifted from Capuno [2001].

13

its higher average score in DOI (47.2 compared to Bulacan’s 44.8) which measures the budget for health, education, nutrition, and other poverty-reducing expenditures. With better provision, it can therefore be expected that the people in Davao del Norte will have a relatively higher level of satisfaction with their local governments, as suggested by the province’s relatively superior average score in DNI (49, compared to Bulacan’s 42).

The wider extent of democratic participation in governance in Davao del Norte is more perhaps due more to the province’s relatively lower population density, which facilitates the closer interaction between local governments and their constituencies. Another critical factor can be the province’s relative isolation from Metro Manila, which, although economically disadvantageous, may be politically beneficial since local governance issues are not as overshadowed by national concerns as in Bulacan. Hence, the local governments in Davao del Norte are able to capitalize on their supposed economic disadvantage to provide a better quality of local governance to their constituencies.

14

Table 2. Scores for 2001 in the GOFORDEV Index GOFORDEV Index Overall Pilot Areas

Development Needs Index Family Condition

Public Problem

Day Care

Health Service

Drinking Water

Dev’t. Orien. Index Total

Development Expenditure Priorities

Participatory Development Index School Board

Local Dev’t. Coun.

BULACAN 41 46 64 Angat* 10 26 85 32 78 0 0 61 53 70 Baliwag** 10 33 81 48 91 100 100 68 52 94 Guiguinto* 19 52 83 51 54 100 100 52 53 26 SJDM City** 25 33 70 50 85 100 100 47 58 59 Plaridel** 23 49 90 52 77 0 0 44 55 27 Bustos* 26 46 83 43 77 100 0 DAVAO DEL NORTE 79 64 78 B.E. Dujali* 12 58 99 93 57 100 100 58 46 56 Panabo City** 10 29 95 15 79 100 100 60 57 28 IGACOS* 4 57 94 77 54 100 100 44 58 36 Sto. Tomas** 15 43 97 68 65 0 0 52 43 43 Tagum City** 7 20 81 54 53 100 100 57 53 29 Asuncion* 12 32 98 73 52 100 100 Note: * Relatively low evel of socioeconomic development ; ** Relatively high level of socioeconomic development.

15

Brgy. Consul tation

Brgy. Meetings

Total

39 32 20 74 76 69

15 5 11 32 17 33

14 59 58 77 23 50

89 61 99 89 62 88

91 22 86 61 22 64

95 71 96 38 71 88

Sensitivity tests The changes in the scores and the relative rankings of the pilot test areas after adjusting the weights applied to the component indicators and sub-indices would indicate whether the GOFORDEV Index formula unduly biases the overall results. Two types of weight adjustments are made. The first type concerns the application uniform weights for each of the component indicators (i.e., each component indicators would be weighted by 0.10). The new weights are significantly different from the current implied weights of each of the component indicators of the DNI (0.066), DOI (0.333) and PDI (0.083).

The second type concerns the increase in the weight of one of the subindices and, correspondingly, a decrease in the weights of the remaining subindices. It can be seen from Table 1 that the three sub-indices are weighted equally (0.0333) in the present formula. Under the second adjustment scheme, the new GOFORDEV Index may then be called “needs-dominant”, “expendituredominant” or “participation-dominant”, if the DNI, DOI or the PDI is given the biggest weight (i.e., depending on one’s view which aspect is most important).

In Table 3, the new scores obtained under the revised formula for the GOFORDEV Index are presented. As expected, the new scores differ from earlier estimates (shown in Table 2). The changes however go in both directions and do not seem to be correlated with the relative levels of socioeconomic development. These imply that the GOFORDEV Index does not induce a bias for or against any locality on the basis of fiscal revenues or resources. This observation is corroborated by the results shown in Table 4 where it can be seen that the relative rankings are not completely reversed after the weight adjustments. Much of the changes in relative rankings appear to be limited to the pilot test areas in Bulacan, whose range of scores in the DOI are much wider than those of the other pilot test areas.

16

Table 3. Scores in the GOFORDEV Index with adjusted weight factors

Pilot test area

BULACAN Angat Baliwag Guiguinto SJDM City Plaridel Bustos DAVAO DEL NORTE B.E. Dujali Panabo City IGACOS Sto. Tomas Tagum City Asuncion

Estimated GOFORDEV INDEX Needs Expenditure Dominant Dominant (DNI=0.5; (DNI=0.25; DOI=0.25; DOI=0.5; PDI=0.25) PDI=0.25)

Actual GOFORDEV Index

Equal weights applied to the component indicators

41 61 68 52 47 44

34.9 57.0 58.4 59.5 44.3 50.4

42.5 58.6 63.8 51.9 49.7 46.8

34.3 60.3 65.3 57.9 40.9 45.5

34.3 60.3 65.3 57.9 40.9 45.5

79 58 60 44 52 57

77.7 56.7 69.9 47.4 54.2 64.8

75.2 54.5 59.7 47.2 50.0 56.0

78.7 57.1 52.4 41.8 50.0 49.9

83.0 60.8 69.4 42.2 57.0 64.6

Participation Dominant (DNI=0.25; DOI=0.25; PDI=0.5)

Table 4. Rankings in the GOFORDEV Index with adjusted weight factors

Pilot test area

BULACAN Guiguinto Baliwag SJDM City Plaridel Bustos Angat DAVAO DEL NORTE B.E. Dujali IGACOS Panabo City Asuncion Tagum City Sto. Tomas

Estimated GOFORDEV INDEX Needs Expenditure Dominant Dominant (DNI=0.5; (DNI=0.25; DOI=0.25; DOI=0.5; PDI=0.25) PDI=0.25)

Actual GOFORDEV Index

Equal weights applied to the component indicators

1 2 3 4 5 6

2 3 1 5 4 6

1 2 3 4 5 6

1 2 5 3 6 4

1 2 3 5 4 6

1 2 3 4 5 6

1 2 4 3 5 6

1 2 4 3 5 6

1 3 2 5 4 6

1 2 4 3 5 6

17

Participation Dominant (DNI=0.25; DOI=0.25; PDI=0.5)

Correlation with other indicators The four panels in Figure 1 show the scatter plots of the GOFORDEV Index with each of the four indicators mentioned above, namely the membership in local organization ratio, trust in elected officials ratio, net satisfaction rating of the mayor, and desirable changes in public services ratio. In each of the four panels, a line is drawn that depicts the underlying relationship between the GOFORDEV Index and one of the four indicators based on the cross-median of the variables in the two axes. The line, which is fairly resistant to outliers, therefore provides the “best fit” of the data, something that cannot be done with the usual regression techniques because of the small number of data points.

On the whole, a positive correlation between the Index and each of the indicators is observed, although the relationship between the Index and either the trust indicator (panel B) or the net satisfaction rating of the mayor (panel C) is best described as U-shaped. The non-linear relationship can perhaps be explained by a number of factors, including the critical role of the local political condition (as in Angat and Sto. Tomas), the reportedly poor visibility of the mayor to his constituents (San Jose del Monte City), and the possible measurement errors (in both the Index and the other four indicators used here). In a sense, therefore, the GOFORDEV Index, despite its simplicity, correlates well with other aspects of governance it does not measure.

18

Figure 2. Scatter plot of the GOFORDEV Index and Participation in Organizations Dujali

GOFORDEV Index

79

Guiguint

Baliwag

IGACOS Panabo

Asuncion SJDM

Tagum

Plaridel Bustos

Sto.Toma

Angat

41 5

53 Participation in Organizations

Figure 3. Scatter plot of the GOFORDEV Index and Trust in Elected Officials Dujali

GOFORDEV Index

79

Guiguint

Baliwag

IGACOS Panabo

Asuncion TSaJgDuM m Plaridel

BSut sot.oTso m a

41

Angat

56

27 Trust-Elected Officials

19

Figure 4. Scatter plot of the GOFORDEV Index and Net Satisfaction with Mayor Dujali

GOFORDEV Index

79

Guiguint

Baliwag

IGACOS Panabo Asuncion

Tagum

SJDM

Plaride S t o . T o mBau s t o s

41

Angat

80

33 Net Satisfaction with Mayor

Figure 5. Scatter plots of the GOFORDEV Index and Desirable Changes in Public Services

Dujali

GOFORDEV Index

79

Guiguint

Baliwag

IGACOS

Panabo

Asuncio

SJDM

Tagum Plaridel

Bustos

41

Sto.Toma

Angat

51

99 Desirable Changes in Public Service

20

5. Concluding remarks

This paper has no more than presented some of the emerging results of the first year of the pilot test of the GOFORDEV Index. Nonetheless, some tentative conclusions about the validity of the Index may be said. In general, the relative rankings of the pilot test areas do not appear to be overly sensitive to the current formula of the GOFORDEV Index, thus indicating some degree of immunity of the Index to possible and uncontrolled measurement errors. Moreover, the Index seems to be consistent with other indicators, thus suggesting that the analysis of the local governance based on the Index will not be qualitatively different from the analyses based on other indices.

These results however should be evaluated further as more data and information are generated during the second-year of the pilot test and subjected to more rigorous statistical tests. In particular, tests that are more sensitive to small variations in the scores are required to firmly establish the internal consistency of the component indicators towards making the Index even more parsimonious.

Additionally, an improved sampling design (more households and areas to cover, better questionnaire) should help reduce the margin of error (10 percent) obtained with only a 100 respondents. Finally, the impact of the Index on local budget decisions and processes and eventually on the people’s welfare must be verified. These will be the subjects of the next report on the validity and usefulness of the GOFORDEV Index.

21

References Bautista, V. A. [1999] Combating Poverty through the Comprehensive and Integrated Delivery of Social Services. U. P. National College of Public Administration and Governance. Quezon City. Bautista, V. A. and L. J. Juan [2000]. “An analysis of the Minimum Basic Needs Approach and potential for assessing governance.” A report submitted to the Philippine Center for Policy Studies. Quezon City. Brillantes, A. [1999]. "Developing Indictors for Good Governance". Draft. U. P. National College of Public Administration and Governance. Quezon City. Brillantes, A. et al. [2000]. Development of Good Governance Indicators for Poverty Programme Assessment. Ateneo de Manila University - Rockwell Campus: Makati City. Burton, E. M. [2000]. “A Baseline study on the Indicators of Good Governance in Davao del Norte.” A report submitted to the Philippine Center for Policy Studies. Quezon City. Capuno, J. J. [2000]. "GOFORDEV Index: Advocating good governance for local development ", Issues & Letters 9 (5 and 6): 1-10. Capuno, J. J. [2001]. “An analysis of the baseline results of the GOFORDEV Index.” Draft. Philippine Center for Policy Studies. Quezon City. Capuno, J. J., M. M. S. Garcia and J. S. Sardalla [2001]. “Promoting local development through good governance: A partial assessment of the GOFORDEV Index”, Issues & Letters 10(3): 1-10. Estrella, M. [2000]. “Review of Literature on Indicators of Good Local Governance.” A report submitted to the Philippine Center for Policy Studies. Quezon City Gopakumar, K. [1998]. "Citizen feedback surveys to highlight corruption in public services. The experience of Public Affairs Centre, Bangalore". Transparency International. Human Development Network and the United Nations Development Programme [1994]. Philippine Human Development Report 1994. Makati City. Human Development Network and the United Nations Development Programme [1997]. Philippine Human Development Report 1997. Makati City.

22

Leitmann, J. [1999]. "Can city QOL indicators be objective and relevant? Towards a participatory tool for sustaining urban development" Local Environment 4 (2): 169-171. Lim, J., J. Doñgail and R. Oliveros [2000]. “Documentation and analysis of experiences of NGOs, social development agencies and people's organizations with local governments.” A report submitted to the Philippine Center for Policy Studies. Quezon City. Local Government Development Foundation [1999] "Performance indicators and service delivery measurement". Draft. Manila City. Manasan, R. G. et al. [1999]. Towards better government. Developing indicators of good governance for local government. National Economic and Development Authority and the United Nations Development Programme. Pasig City. National Economic and Development Authority [2001]. Medium Term Philippine Development Plan 1999-2004. Pasig City. Philippine Business for Social Progress [2001]. A compilation of indicators of good governance. A report prepared for the Donors Forum on LGU Benchmarking. Manila. Pimentel, A. Q. Jr. [1993]. The Local Government Code of 1993. The Key to National Development. Mandaluyong City: Cacho Publishing House, Inc. Pingol, A. and V. Alonzo [2000]. “Baseline study of Ilocos Norte.” A report submitted to the Philippine Center for Policy Studies. Quezon City. Rocamora, J. and R. Villamil [2000]. “Siquijor Socio-economic and Political Profile.” A report submitted to the Philippine Center for Policy Studies. Quezon City. Root, H. [1995]. " Managing Development through Institution Building.” Occasional Papers No. 12. Asian Development Bank. Mandaluyong City. Solon, O., R. Fabella and J. Capuno [2001]. “Is local development good politics? Local development expenditures and the re-election of governors in the Philippines for 1992, 1995 and 1998.” A report submitted to the Philippine Center for Policy Studies. Quezon City. Veneracion, J. B. [2000]. “A Baseline Study of Bulakan.” A paper submitted to the Philippine Center for Policy Studies. Quezon City. World Bank [2000]. Philippines. Growth with equity: the remaining agenda. Washington, D.C. 23

World Bank [2001]. Philippines. Filipino Report Card on pro-poor services. Washington, D.C.

24